Adolescents, especially college students, are at high risk for problem drinking, but the resulting long term consequences on brain development and function are not well studied. In this 5-year proposal, the investigators will recruit a total of 2000 18-21 year old first-year students (50% male) from 2 demographically distinct local colleges in CT; Central Connecticut State University and Trinity College. All students will be characterized at baseline at their respective campuses on a comprehensive battery of cognitive tasks, some assessing risk for problem drinking in the context of an integrative risk model and others previously identified as being sensitive to effects of alcohol abuse in adolescents. All participants will also be assessed on instruments measuring amount and patterns of current and past alcohol and drug use, consequences of use (eg mood alteration, DWI), and family history of substance use. Academic grades and SAT/ACT scores will be recorded. The sample will be stratified by alcohol use patterns and a total of 300 representative members derived from use quartiles (over-represented from the heaviest quartile, especially binge-pattern users) will undergo additional detailed assessment at the Olin Neuropsychiatry Research Center on 4 functional MRI (fMRI) paradigms assessing a. hippocampal integrity (a virtual maze spatial memory task), b. working memory circuitry (a modified Sternberg paradigm), c. reward anticipation (a monetary incentive delay task) and d. ability to inhibit a pre-potent stimulus (a Go/No-Go task), plus structural MRI and urine toxicology testing. Monthly, all participants will complete web-based ratings of alcohol and drug use amounts and patterns, mood, life events, stress and adverse consequences of use. Two years post initial testing, all available participants (~90%) will repeat the initial cognitive test battery (using alternate test forms) and detailed substance use measures. Of the 300 initial fMRI participants, the ~85% available and willing will repeat the imaging paradigms. The 3 overarching aims are to identify initial predictors of persisting alcohol and drug abuse, to detect early evidence of alcohol and drug-related impairment of brain function & cognitive problems and to determine whether particular patterns of drinking, especially binging, are most associated with cognitive decline over the 2 yrs of assessment. As part of the project we will test a theoretical model of how brain structure and function explain key relationships between risk and alcohol/substance use outcome, incorporating highly novel measures using data fusion, in order to better accomplish the overall goals. The PI and his collaborators have extensive experience with the tasks and procedures used in this Project, including web-based tracking of alcohol use. PUBLIC HEALTH RELEVANCE: US college students are at high risk for problem drinking. This project will recruit and test cognitively 2000 first-year students from local colleges in Connecticut, fMRI scan a representative sub-sample and assess alcohol and drug use by web-based reporting over 2 years, when all students will be retested/rescanned. Findings from this study will provide important insights on risk factors for problem drinking and how alcohol use impacts the developing adolescent brain.